Starting Data: Philadelphia
This was the first data we considered. We, like the article, moved first to consider the number of complaints for each officer. However, there is a lot of other detail in these data sources that can be considered.
Similar Class of Top of Offenders in All Three Cities
Deciles were calculated by placing an officer into one of ten equal groups of officers in his city based on his amount of complaints. Those in decile 10 were in the top 10% of amount of complaints about them.
It is interesting not only that there is a class of unusually high offenders, but also that in every city this class contributes roughly the same proportion of complaints, 1/3. Are there other similarities between these groups in each city besides how many complaints they earn?
Plots
The class made several nice renditions of the data above, including Sophia, who has already started working on a policy proposal. The numbers in our analysis differ from those presented in the FT Article, perhaps because they were more discriminate against investigations that have not yet been closed.
Race
After briefly touching on race and policing earlier in the article, the article moved to discuss and show an interesting graphic on social groups and number of complaints. The following graph and table considers the racial composition of complaint deciles.
Unlike in the plot, racial composition here is percent members of race in a decile.
There remain many questions that could be explored on this data. First on my mind: how many officers are excluded from this data? i.e. How many officers have never recieved a complaint? Alina’s dogged pursuit of an answer to her own question on the Discord this week answered a question I had had for a long time.
If you feel this list is incomplete, reach out to one of the course staff.